Investigation of the performance and interpretability of two models, a large language models (LLM) and a small-scale model, trained on low-resource language pairs Xhosa Zulu and Tswana-Zulu
This submission contains images and datasets used in the research for a dissertation "Assessing interpretability in machine translation models for low-resource languages".
The images include machine translation model-generated heatmaps and machine translation model-generated translations.
The datasets include the following:
BLEU scores from model training and graphs
Post model evaluation results for MQM and graphs
Post model evaluation results for ESS and results
Small-scale model training results comparisons with generated graphs [to evaluate early stopping]
History
Department/Unit
Engineering, Built Environment and Information Technology